Probabilistic Methods in Cancer Biology

Recent advances in experimental techniques have made it possible to generate an enormous amount of ‘raw’ biological data, with cancer biology being no exception. The main challenge faced by cancer biologists now is the generation of plausible hypotheses that can be evaluated against available data and/or validated through further experimentation. For persons trained in control theory, there is now a significant opportunity to work with biologists to create a virtuous cycle of hypothesis generation and experimental validation. Given the large number of uncertain factors in any biological experiment, probabilistic methods are natural in this setting. In this paper, we discuss four specific problems in cancer biology that are amenable to study using probabilistic methods, namely: reverse engineering gene regulatory networks, constructing context-specific gene regulatory networks, analyzing the significance of expression levels for collections of genes, and discriminating between drivers (mutations that cause cancer) and passengers (mutations that are caused by cancer or have no impact). Some research problems that merit the attention of the controls community are also suggested.